Backlash on Fineline R&P

[Sheridan Basses]

Hi All. I'm getting errors of approx .065-070 on my x-axis. The motors are tuned to the appropriate steps, 1909.861. For the life of me, I cant get this to cut accurately without compensation. It is slightly better with a smaller diameter bit, but is excessive regardless. Any ideas? Runout appears to be minimal with a mic, and I can tell it to jog 2 inches and it will accurately do that. But once I turn it on the router (porter cable), the accuracy drops on the x.

[jsheerin]

If it's accurate for one condition and not for another (router running), the problem is likely due to something caused by the router. Could be electrical interference, missed steps from the motors under the load of cutting, or several other things.

However you haven't really described backlash. When you say you're getting an 'error' what exactly does it do?

If it only does it with the router running, does it do it when the router is running but you aren't cutting any material? Checking this would let you try to isolate it to an electrical interference problem or mechanical / motor strength problem.

What are your cutting parameters (speed, acceleration, depth of cut, size of bit, material being cut)?

[Sheridan Basses]

Hi. So here's what I've been able to observe.
- Using a 3/8 end mill.
- Pine material
- If I place a ruler down and jog the x 2 inches (with router off), it moves exactly 2 inches. I've double checked this by inserting a .040 bit and observing the same movement.
- My test has been 2.5 x 2.5 inch square pocket in pine. The X cut is always short by .065-070 on the x. I checked with a 1/4 bit and it was the same, give or take a few thousands. Giving the x .070 of backlash compensation brings it to the desired width of 2.5 inches

Speed is 21000 for the spindle. (It's fixed speed Porter Cable)
Cut depth - 1/8
Feed rate is 80%
x-velocity is 43.1, Accel is 10.2
y velocity is 69, Accel is 16

I'll try running the ruler test with the router spinning. It seems logical that the bit is encountering a .070 deflection under load, but I'm not sure. I can look into borrowing another router, but that's a few days from now. Can you suggest anything else i should try in the meantime?
Thanks

[jsheerin]

Okay, if backlash compensation fixes the problem, I can see why you would think it is that. However from what you're saying it sounds like you only see it when the machine is under load.

Can you post some pictures of your machine, especially the x drive mechanism? Maybe we can see something off... Also, check that any set screws are tightened on the x mechanism.

Do you notice that only one side of the gantry gets off, or do both sides? You could try cutting the pocket at different sides of the table to check this. If it's one side, I'd say it's more likely to be mechanical - maybe something loose. If it's both sides doing something exactly the same, I'd say it's more likely to be electrical.

To clarify, is your machine moving at 43.1 inches per minute while cutting? If so, try running the same test at a slower speed, like 10ipm. However if you're using the Fineline setup, the speeds and accelerations you're running should already be pretty conservative...

Can you feel any deflection if you push / pull on the router in the x direction with your hand?